Disruption of a single copy of the SERCA2 gene results in altered Ca2+ homeostasis and cardiomyocyte function.

A mouse model carrying a null mutation in one copy of the sarcoplasmic reticulum (SR) Ca(2+)-ATPase isoform 2 (SERCA2) gene, in which SERCA2 protein levels are reduced by approximately 35%, was used to investigate the effects of decreased SERCA2 level on intracellular Ca(2+) homeostasis and contractile properties in isolated cardiomyocytes. When compared with wild-type controls, SR Ca(2+) stores and Ca(2+) release in myocytes of SERCA2 heterozygous mice were decreased by approximately 40-60% and approximately 30-40%, respectively, and the rate of myocyte shortening and relengthening were each decreased by approximately 40%. However, the rate of Ca(2+) transient decline (tau) was not altered significantly, suggesting that compensation was occurring in the removal of Ca(2+) from the cytosol. Phospholamban, which inhibits SERCA2, was decreased by approximately 40% in heterozygous hearts, and basal phosphorylation of Ser-16 and Thr-17, which relieves the inhibition, was increased approximately 2- and 2.1-fold. These results indicate that reduced expression and increased phosphorylation of phospholamban provides compensation for decreased SERCA2 protein levels in heterozygous heart. Furthermore, both expression and current density of the sarcolemmal Na(+)-Ca(2+) exchanger were up-regulated. These results demonstrate that a decrease in SERCA2 levels can directly modify intracellular Ca(2+) homeostasis and myocyte contractility. However, the resulting deficit is partially compensated by alterations in phospholamban/SERCA2 interactions and by up-regulation of the Na(+)-Ca(2+) exchanger.

In heart, muscle relaxation is largely dependent on the action of the sarcoplasmic reticulum (SR) 1 Ca 2ϩ ATPase (SERCA) to resequester cytosolic calcium released during contraction. Increased activity of SERCA, either by transgenic overexpression of SERCA isoforms in the heart (1)(2)(3) or by ablation of its regulatory protein, phospholamban (PLB) (4), has been shown to enhance cardiac rates of contraction and relaxation (1)(2)(3)(4). To examine the effects of decreased SERCA2 activity on cardiac function, we have recently developed a transgenic mouse model with a null allele of the SERCA2 gene (5). Although complete loss of SERCA function in homozygous animals is embryonic lethal, disruption of one copy of the SERCA2 gene results in decreased cardiac SERCA2 mRNA (ϳ45%), protein (ϳ35%), and SR Ca 2ϩ uptake (ϳ35%) (5). These changes are associated in vivo with impaired cardiac performance (5). Because SERCA2 activity controls both the rate of calcium removal and the amount of calcium stores available within the SR, we hypothesize that the level of SERCA2 activity is a critical determinant of cardiac contractility. Therefore, one goal of this study is to determine if reduced SERCA2 levels compromise cardiac contractility by directly altering calcium handling and contractile functions of individual myocytes during excitationcontraction coupling.
During excitation-contraction coupling, Ca 2ϩ entry through the L-type Ca 2ϩ channel activates Ca 2ϩ release from SR Ca 2ϩ stores, via the ryanodine receptor (RyR). This rise in cytosolic Ca 2ϩ initiates muscle contraction. In the relaxation phases most of the released Ca 2ϩ is subsequently resequestered into the SR by SERCA, the activity of which is closely regulated by its interaction with PLB. In addition, the released Ca 2ϩ is also extruded via the Na ϩ -Ca 2ϩ exchanger (NCX) and the plasma membrane Ca 2ϩ -ATPase (6). We hypothesize that chronic loss of SERCA2 in SERCA2 heterozygous hearts may be partially compensated by altered expression of other proteins involved in calcium homeostasis during muscle contraction and relaxation. To test this hypothesis, we have examined expression of several proteins known to be important in the control of calcium transients in the heart. Alternatively, decreased contractility in heterozygous hearts may be associated with changes in the expression of contractile proteins, such as switching between ␣and ␤-myosin heavy chain (MHC) isoforms, which occurs during cardiac hypertrophy (7). Therefore, we also examined the expression of myosin heavy chain isoforms in SERCA2-deficient hearts.
We have previously demonstrated that a decreased SERCA2 level in heterozygous hearts results in impaired cardiac function in vivo (5). In this study, we show that, despite multiple changes in expression and function of other calcium-handling proteins (such as PLB, triadin, and NCX), peak calcium transients and SR calcium stores of SERCA2-deficient myocytes are significantly diminished and contractile function is impaired. These data suggest that changes in other calcium-handling proteins are functionally insufficient to completely compensate for the loss of the calcium pump protein and that the level of SERCA protein is an important critical regulator of cardiac function.

EXPERIMENTAL PROCEDURES
As described previously, gene targeting was used to delete part of the promoter, the transcript initiation site, and the first two coding exons of the SERCA2 gene (5). Heterozygous mice, containing a single functional allele of the SERCA2 gene, were identified by polymerase chain reaction genotyping and used in the following studies at 12-14 weeks of age.
Ribonuclease Protection Assays-The riboprobes for mouse cardiac RyR, calsequestrin, triadin, NCX, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were generated by subcloning relevant fragments of genomic DNA or cDNA into pBluescript SK ϩ . These fragments included: mouse cardiac ryanodine receptor gene (RyR2) corresponding to nucleotides ϩ221 to ϩ400 (a 180-base pair (bp) fragment) from the 3Ј-untranslated region (8); mouse cardiac calsequestrin cDNA (a gift from Dr. Evangelia Kranias) (9), corresponding to nucleotides ϩ421 bp to ϩ570 bp (150 bp) relative to the start codon; mouse triadin gene isoform 3, corresponding to nucleotides ϩ22 bp to ϩ 170 bp (150 bp) obtained from GenBank (accession number AA611321); mouse Na ϩ -Ca 2ϩ exchanger isoform 1, consisting of a fragment Ϫ31 bp to ϩ118 bp (150 bp) relative to the AUG codon (10); and mouse GAPDH gene (sequence obtained from GenBank (accession number M32599)), corresponding to the region from Ϫ46 bp to ϩ179 bp (225 bp) relative to the AUG codon. The MAXI script in vitro transcription kit (Ambion, Inc.) was used to synthesize 32 P-labeled cRNA probes from NotI-linearized plasmid templates. Total RNA was isolated from hearts using the Ultraspec-II RNA isolation system (Biotex Laboratories, Houston, TX). 5 g of total RNA (n ϭ 6 for each genotype) was hybridized with the riboprobes described above and processed using RPA III ribonuclease protection assay kit (Ambion) (5,10). The protected fragments were separated by electrophoresis in a 5% denaturing polyacrylamide gel and analyzed by autoradiography. PhosphorImager and ImageQuaNT software (Molecular Dynamics, Wayzata, MN) were used for quantitation of mRNA levels relative to GAPDH loading controls.
To determine the basal phosphorylation level of PLB, polyclonal antibodies raised against a PLB peptide phosphorylated at Ser-16 (PLB-phosphoserine 16) or at Thr-17 (PLB-phosphothreonine 17) (Fluo-reScience Ltd., Reeds, UK) (15) were used. The mouse hearts were excised and immediately freeze-clamped using instruments precooled with liquid nitrogen to ensure complete and rapid freezing of the cardiac tissues. Cardiac homogenates were prepared as described (12,13). The homogenate buffer contained (in mM): imidazole (pH 7.0) 10, sucrose 300, and EDTA 1. The buffer was supplemented with the following proteinase inhibitors: 0.3 mM phenylmethylsulfonyl fluoride, 0.5 mM dithiothreitol, 2 g/l pepstatin A, 10 g/l leupeptin, and 10 g/l aprotinin (13), and 25 mM sodium fluoride was added as phosphatase inhibitor. After separation on 15% SDS-polyacrylamide gels, the homogenate protein was transferred onto a PVDF membrane. The membranes were incubated with PLB-phosphoserine 16 (1:10,000 dilution) and PLB-phosphothreonine 17 (1:5000 dilution) antibodies. ECL and quantitation of the signals were carried out as described above. Both the high molecular weight form of PLB (PLB H ) and the low molecular weight form of PLB (PLB L ) were quantified, and the ratio of PLB H to PLB L was measured. To measure the relative ratio of PLB to SERCA protein, the signals of PLB H and SERCA2a from the same membrane were quantitated.
[ 3 H]Ryanodine Binding Assay-Ryanodine radioligand binding assays were performed as described (16,17). The binding medium contained (in mM): HEPES 20 (pH 7.1), and KCl 600, with free [Ca 2ϩ ] adjusted to a final concentration of 20 M by appropriate addition of 0.5 mM EGTA using the computer algorithms of Robertson and Potter (18). Cardiac homogenates in assay buffer (100 g of total protein) with various concentrations (0.  (2,19,20). Half of the cells from each heart were then used for mechanical studies, and the other half were used for measurements of intracellular free Ca 2ϩ transient. Intracellular calcium transients were measured as described previously (2,19,20). Briefly, myocytes were loaded with 7.5 M Fura-2 AM at 37°C for 15 min in the dark. After loading, cells were washed and resuspended in Ca 2ϩ free Tyrode's buffer. Cytosolic free calcium was measured by ratio imaging of 340-to 380-nm excitation fluorescence of Fura-2 AM (emission wavelength, 510 nm), using a photoscan dual spectrophotometer (Photon Tech, Inc., Santa Clara, CA) coupled to an Olympus IMT-2 UV fluorescence microscope with UV-transparent optics. Cells were selected for use if they were quiescent when unstimulated, and contracted robustly upon field stimulation. Cells were perfused with Sanguinetti solution (in mM: NaCl 112, KCl 5, CaCl 2 0.5, MgSO 4 1.2, NaH 2 PO 2, NaHCO 3 28, glucose 10), bubbled with 95% O 2 /5% CO 2 , and field-stimulated at 0.25 Hz (Grass SD9 stimulator) until twitch characteristics were repeatable, indicating stable Ca 2ϩ loading of the SR. Following field stimulation protocols, cells were exposed to ionomycin, EGTA, and Mn 2ϩ to determine background fluorescence and dye compartmentalization. The intracellular Ca 2ϩ kinetics were analyzed using Origin 4.1 (Microcal Software).
The mechanical properties of myocytes were measured as described previously (2,19,20). Briefly, cells were placed in a well on the stage of an inverted microscope and perfused continuously with oxygenated physiological buffer. Myocytes were field-stimulated at 0.25 Hz, or at 0.5 Hz for at least 40 s per pacing rate. Cell images were videotaped, and myocyte length, percentage of shortening, and peak rates of shortening (ϩdL/dt) and relengthening (ϪdL/dt) were quantitated by comparison with a calibrated micrometer on the microscope stage. At least five cells were examined per mouse, and the values were averaged for mechanical parameters and Ca 2ϩ kinetics. Statistical analysis was based on the number of animals rather than the number of cells.
Assessment of SR Ca 2ϩ Load in Isolated Myocytes-A solenoid perfusion system (W. Barry Co., Boulder, CO) was used to rapidly apply 10 mM caffeine to induce release of SR Ca 2ϩ and assess the SR Ca 2ϩ load (21,22). Cells were perfused with Sanguinetti solution and field-stimulated at 0.25 Hz until twitch characteristics stabilized before each caffeine application. Caffeine was then applied for 10 s. The amplitude of the caffeine-induced Ca 2ϩ transient was used as an index of SR Ca 2ϩ content (23,24).
Measurement of Na ϩ -Ca 2ϩ Exchanger Current Density-Na ϩ -Ca 2ϩ exchanger currents were recorded using whole-cell patch-clamp techniques as described previously (25)(26)(27). Briefly, isolated cardiomyocytes were prepared, and Na ϩ -Ca 2ϩ exchanger currents were recorded at room temperature. External solution contained (in mM): NaCl 135, CaCl 2 2, CsCl 10, MgCl 2 1, glucose 10, HEPES 10, pH 7.3, and pipette solution contained (in mM): CsCl 135, MgCl 2 2, NaCl 15, 1,2-bis(2aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid (BAPTA; Molecular probes) 0.05, and HEPES 10 (pH 7.2). Ryanodine (0.01 mM) and nifedipine (0.01 mM) were included in the external solution to block SR Ca 2ϩ release and Ca 2ϩ entry through the L-type Ca 2ϩ channel. The Ni 2ϩ -sensitive current was obtained by subtracting current records obtained during exposure to 5 mM Ni 2ϩ from the control record, which was measured in the absence of Ni 2ϩ . Recordings were done using an Axopatch 200B patch clamp amplifier (Axon Instruments, Foster City, CA), interfaced to a personal computer. Data were stored in the computer for analysis using custom-written software.
Statistical Analyses-Results are expressed as mean Ϯ S.E. and statistically evaluated by the ANOVA test followed by the Student t test. p Ͻ 0.05 was considered to be the threshold for statistical significant.

Cardiomyocytes from SERCA2 Heterozygous Hearts Have
Lower Cytosolic Peak Ca 2ϩ Transients-We have recently shown that disruption of one copy of the SERCA2 gene results in decreased SERCA2 mRNA (ϳ45%), protein (ϳ35%), and SR Ca 2ϩ uptake (ϳ35%) and that these changes are associated in vivo with impaired cardiac performance (5).
To determine whether this decrease in SERCA pump level affects myocyte Ca 2ϩ homeostasis on a beat-to-beat basis, cytosolic Ca 2ϩ transients were measured. Left ventricular myocytes isolated from wild-type and heterozygous hearts were loaded with Fura-2 AM, and the Ca 2ϩ signals during electrical pacing at 0.25 Hz were measured. The baseline cytosolic calcium level was not significantly different between wild-type and heterozygous cells. But the peak amplitudes of calcium transients were decreased by ϳ30 -40% in heterozygous myocytes (Fig. 1A, Table I), indicating that the absolute amount of Ca 2ϩ available for myofibrillar contractility was decreased. However, the rate of Ca 2ϩ decline as fitted by exponential decay () in heterozygous myocytes was not different from that of wild-type myocytes (Fig. 1A, Table I), suggesting no significant difference in the rate of Ca 2ϩ removal from the cytosol.
SR Ca 2ϩ Load Is Significantly Decreased in SERCA2 Heterozygous Hearts-Because SERCA2 heterozygous myocytes showed a significant decrease in Ca 2ϩ amplitude upon twitch stimulation, we next determined whether the decrease in Ca 2ϩ amplitude was due to a decrease in SR Ca 2ϩ stores or due to a decrease in Ca 2ϩ efflux from the SR. Fura-2 AM-loaded myocytes were exposed to caffeine, and peak Ca 2ϩ transients were measured. We found that amplitudes of the Ca 2ϩ transients in the presence of caffeine were decreased ϳ40 -60% in heterozygous myocytes relative to the transients observed in wild-type cells (Table I). Because caffeine binding opens the RyR, causing Ca 2ϩ release from the SR, the amplitude of the caffeine-induced Ca 2ϩ transient can be used as an index of SR Ca 2ϩ content (23,24). Our data show that reduced SERCA activity leads to significantly decreased RyR-releasable SR Ca 2ϩ stores in heterozygous myocytes.
Decreased Contractility of SERCA2 Heterozygous Myocytes-To determine whether the observed changes in SR Ca 2ϩ transport affect cardiac contractility, mechanical properties of isolated cardiomyocytes were examined. There was no significant difference in the resting cell length between wild-type and heterozygous myocytes. However, there was ϳ30% decrease in the percentage of cell shortening in heterozygous myocytes (Fig. 1B, Table I). In addition, the rate of cell shortening (ϩdL/ dt), as well as the rate of cell relengthening (ϪdL/dt) was significantly decreased (ϳ 40%) in heterozygous myocytes (Fig.  1B, Table I). These data demonstrate that a reduction in SERCA2 results in decreased cytosolic Ca 2ϩ transients and SR Ca 2ϩ load and that these alterations can directly affect contractile function of the individual myocyte.
Expression Levels of Myosin Heavy Chain Isoforms Are Unchanged in SERCA2 Heterozygous Hearts-Alterations in the expression levels of myosin heavy chain isoforms ␣and ␤-MHC can be an important determinant of the dynamic function of the myocardium and are correlated with the maximum velocity of muscle shortening (28,29). Therefore, it can be argued that the impaired myocyte contraction and relaxation in heterozygous hearts is also due to a switch in myosin heavy chain isoform expression. In addition, induction of ␤-MHC has been reported to be a marker of cardiac hypertrophy (7), a condition that is also associated with decreased levels of SERCA pump expression (30). In this study, we examined the expression levels of ␣-MHC and ␤-MHC in SERCA2 heterozygous hearts. Our results showed that the expression levels of both isoforms were unchanged (␣-MHC, 107 Ϯ 2%; ␤-MHC, 107 Ϯ 1% in heterozygous hearts compared with wild-type controls, 100%; n ϭ 4 of each genotype) (Fig. 2). These results indicate that the impaired contractile function in heterozygous myocyte is not due to a change in myosin heavy chain isoforms. Gravimetric analysis showed similar heart:body-weight ratios between wild-type and heterozygous hearts (4.10 Ϯ 0.05 mg/g, heterozygous; 4.12 Ϯ 0.54 mg/g, wild-type; n ϭ 6 of each genotype). Therefore, there is no apparent indication of cardiac hypertrophy in SERCA2 heterozygous mice.
The Amount of PLB Protein Is Decreased but Its Phosphorylation Status Is Increased in SERCA2 Heterozygous Mice-Despite the decreased level of SERCA2 expression, the apparent rate of Ca 2ϩ decline () during Ca 2ϩ transients was unaltered (Fig. 1A). This result suggests that the decrease in SERCA2 pump activity might be compensated by SERCA2/PLB interaction in myocytes of heterozygous mice. Recent studies have shown that both the PLB:SERCA2 protein ratio and the PLB phosphorylation status regulate SERCA pump function (4,19,31,32). To determine whether PLB regulation of SERCA activity is altered in SERCA2 heterozygous hearts, we examined the amount of PLB protein, as well as its phosphorylation status, using quantitative immunoblot analysis. To maintain the relative ratios of both high molecular weight and low molecular weight PLB forms, some samples were analyzed with- out boiling (which reduces PLB to monomer) prior to gel electrophoresis. As shown in Fig. 3, A and B, PLB protein (both high molecular weight PLB (PLB H ) and low molecular weight PLB (PLB L )) levels were significantly decreased in SERCA2 heterozygotes (PLB H , 60 Ϯ 7%; PLB L , 45 Ϯ 6%; wild-type, 100%; n ϭ 6 of each genotype). Moreover, the relative ratio of PLB H to PLB L was increased ϳ1.4-fold (0.99 in wild-type; 1.40 in heterozygous; Fig. 3A), suggesting a shift toward the oligomeric form of PLB in heterozygous hearts.
We recently showed that the PLB mRNA level was similar between wild-type and heterozygous hearts (5). However, our protein analyses show that the PLB H protein is decreased and its level is comparable to the decreased SERCA2 protein level (ϳ35%) (5). To confirm these findings, we further quantitated PLB H and SERCA2 protein levels and determined the PLB H : SERCA2 protein ratio. In comparison to wild-type hearts (relative ratio of PLB H to SERCA2 was set as 1.0), the PLB H to SERCA2 ratio was 1.1 Ϯ 0.1 in heterozygous hearts (n ϭ 6 of each genotype, p Ͼ 0.05), indicating no significant change in heterozygous hearts. Thus, our data suggest that the SERCA2 protein level may affect the steady-state level of PLB protein to maintain an optimal PLB:SERCA2 ratio in SERCA2 heterozygous mice.
Interactions between PLB and SERCA2 are also regulated by the phosphorylation status of PLB. Phosphorylation of PLB occurs at two sites: serine 16 (by cAMP-dependent protein kinase) and threonine 17 (by Ca 2ϩ /calmodulin-dependent protein kinase, CaM kinase II) (33,34). To determine whether the phosphorylation status of PLB is altered subsequently to decreased SERCA pump level, we analyzed PLB phosphorylation in vivo under basal conditions using phospho-specific PLB antibodies and Western blot analysis. There is increased basal PLB H phosphorylation at Ser-16 (ϳ200%) and Thr-17 (210%) residues in SERCA2 heterozygous hearts compared with wildtype hearts (set to 100%, for wild-type, n ϭ 6 of each genotype, Fig. 3A). These data were further confirmed by analyzing boiled samples, which reduces PLB to the monomeric form, PLB L (Fig. 3B). When normalized to the decreased PLB H protein level, phosphorylation at PLB H Ser-16 and Thr-17 sites was increased ϳ3.3and ϳ3.5-fold in heterozygous hearts, respectively. In addition, the ratio of both Ser-16 and Thr-17 phosphorylation in the pentameric form PLB H to the monomeric form PLB L were increased ϳ1.4and ϳ1.3-fold, respectively, in heterozygous mice (Fig. 3A). These results suggest that decreased SERCA2 pump activity results in multiple compensatory changes in PLB, including decreased PLB protein level, increased expression of the oligomer form of the protein, and increased basal PLB phosphorylation.
The Na ϩ -Ca 2ϩ Exchanger Protein Expression and Function Are Up-regulated in SERCA2 Heterozygous Hearts-SR Ca 2ϩ uptake is thought to be the dominant mechanism responsible for the rapid decrease in cytosolic free Ca 2ϩ (35,36). However, Ca 2ϩ can also be extruded from the cell via the sarcolemmal Na ϩ -Ca 2ϩ exchanger (NCX) (6,35). Therefore, to determine whether a decreased amount of SERCA2 resulted in compensatory alterations in NCX function, we measured NCX expression levels and determined currents by patch-clamp analysis.
Patch-clamp analysis was used to analyze NCX current in heterozygous myocytes. Ni 2ϩ -sensitive current (which represents the NCX current) was determined by subtracting control current records from those obtained during exposure to 5 mM Ni 2ϩ . As shown in Fig. 5 (A and B), the sustained outward  2. Quantitation of ␣and ␤-myosin heavy chain proteins in wild-type and heterozygous hearts. Four individual hearts were used for each group. 3 g of homogenate protein was subjected to SDS-polyacrylamide gel electrophoresis, transferred to PVDF membranes, and probed with specific antibodies. The signals for ␣and ␤-myosin heavy chain in hearts from wild-type (W) and heterozygous (H) mice were quantitated relative to that of ␣-sarcomeric actin. current was increased by ϳ40% (0.65 Ϯ 0.07 mV in wild-type versus 0.91 Ϯ 0.11 mV in heterozygous, p Ͻ 0.05), and the inward tail current was increased by ϳ43% (0.60 Ϯ 0.07 mV versus 0.86 Ϯ 0.10 mV, p Ͻ 0.05). These data confirm that increased NCX protein levels result in increased functional NCX activity, both in forward and reverse current direction.  4. Expression levels of Na ؉ -Ca 2؉ exchanger in wild-type and heterozygous hearts. A, mRNA levels of Na ϩ -Ca 2ϩ exchanger (NCX) was determined by RNase protection assay. The hybridization signals for NCX mRNA in hearts from wild-type (W) and heterozygous (H) mice were quantitated relative to that of the GAPDH. The bar graph shows the mean Ϯ S.E. of six pairs of wild-type (WT) and heterozygous (HET) hearts. B, representative immunoblot of NCX protein from six wild-type and heterozygous hearts. Increasing amounts (5-20 g) of tissue homogenates from wild-type and heterozygous hearts were separated by SDS-polyacrylamide gel electrophoresis, and probed with specific antibodies. The bar graph shows the mean Ϯ S.E. of six individual hearts. WT, wild-type mice; HET, heterozygous mice. of SR Ca 2ϩ Release and Ca 2ϩ Storage Proteins-To determine whether SERCA2 gene ablation affects the expression of SR Ca 2ϩ release and storage proteins, mRNA levels of RyR, triadin, and calsequestrin were analyzed by RNase protection. As shown in Fig. 6A, the mRNA levels of RyR, triadin, and calsequestrin were not significantly different between wild-type and heterozygous mice. Quantitation of the levels of RyR and calsequestrin protein showed no significant change in SERCA2 heterozygous mice compared with wild-type mice (Fig. 6B). We also performed radioligand receptor binding studies to measure functional ryanodine receptor levels. Scatchard plot analysis showed the B max of the RyR binding was decreased by ϳ6% (105.11 Ϯ 1.89 pmol/mg in wild-type versus 111.51 Ϯ 2.06 pmol/mg in heterozygous, n ϭ 6, p ϭ 0.022), whereas the dissociation constant of RyR binding (K d ) remained unchanged (2.24 Ϯ 0.18 nM, wild-type versus 2.28 Ϯ 0.04 nM, heterozygous; p Ͼ 0.05) (Fig. 7).

SERCA2 Gene Ablation Differentially Alters the Expression
Quantitative immunoblot analysis showed that the amount of triadin isoform 3 (37,38) was increased by ϳ40% (140 Ϯ 3%, heterozygous; with wild-type set to 100%) (Fig. 6B). Although ryanodine receptor and calsequestrin levels were not significantly altered, up-regulation of triadin levels suggest a potential role for this protein in SR Ca 2ϩ release.

SERCA Pump Level Is a Critical Determinant of Both Relaxation and Contraction of Cardiomyocytes-
A major goal of this study was to investigate whether a decrease in SERCA pump level is sufficient to alter intracellular Ca 2ϩ homeostasis (SR Ca 2ϩ stores, Ca 2ϩ release, and Ca 2ϩ removal) and cardiomyocyte contractility. Our data demonstrate that both the peak amplitudes of the cytosolic Ca 2ϩ transients and SR Ca 2ϩ stores (as measured by caffeine-induced Ca 2ϩ release) were decreased by ϳ30 -40%, and ϳ40 -60%, respectively, in myocytes of SERCA2 heterozygous mice. Functional studies at the myocyte level (current study), as well as in intact heart in vivo (5) revealed that not only the rate of relaxation is decreased but also the rate of contraction. Importantly, these changes occur in the absence of a switch from ␣-MHC to ␤-MHC (Fig. 2).
These data, taken together, suggest that a decreased SERCA pump leads to decreased SR Ca 2ϩ uptake function, thereby resulting in decreased SR Ca 2ϩ store and Ca 2ϩ release. This decreased SR Ca 2ϩ release, hence, limits the Ca 2ϩ availability for contractile protein activation. We have recently shown that, in SERCA-overexpressing mice, an increase in SERCA pump level leads to enhanced rates of contraction and relaxation (1)(2)(3). These results demonstrate that the SERCA pump is a critical regulator of both contraction and relaxation of cardiomyocytes.
SERCA/PLB Interaction Is Modified to Compensate for Decreases in SERCA Pump Level and Activity-In this study, we found that the rate of Ca 2ϩ decline () in heterozygous myocytes was not different from that of wild-type myocytes (Fig.  1A, Table I), suggesting that the Ca 2ϩ removal function is compensated in SERCA2 heterozygous hearts. A novel finding of this study is that a decrease in SERCA activity is partially compensated by adjusting PLB expression level. Using Western blot analysis, we showed that PLB H protein levels were decreased by ϳ40% in SERCA2 heterozygous hearts, which is comparable to that of the decrease in SERCA2 pump levels (ϳ35%) (5). Therefore, the PLB:SERCA2 ratio, which has been shown to be one of the main determinants of the affinity of FIG. 5. Na ؉ -Ca 2؉ exchanger currents in wild-type and heterozygous mice. A, representative tracings of Na ϩ -Ca 2ϩ exchange currents from myocytes during 1.5-s depolarizing pulses from a holding potential of Ϫ60 mV to ϩ80 mV. The Ni 2ϩ -sensitive currents (c) were obtained by subtracting current records obtained during exposure to 5 mM Ni 2ϩ (b) from the control record (a). B, Na ϩ -Ca 2ϩ exchanger current density (normalized to the cell capacitance) shows a significant increase in the sustained outward current and peak tail current density of the Ni 2ϩ -sensitive current from six heterozygous (HET) mice as compared with five wild-type (WT) mice. One to three cells were selected from each heart, "n" in parentheses represents the numbers of cells in wildtype and heterozygous hearts. Data represent mean Ϯ S.E. *, p Ͻ 0.05.
FIG. 6. Expression levels of RyR, calsequestrin, and triadin in wild-type and heterozygous hearts. A, RNase protection analysis of RyR, calsequestrin (CSQ), and triadin mRNA. The hybridization signals in hearts from six pairs of wild-type (W) and heterozygous (H) mice were quantitated relative to that of the GAPDH. B, representative Western blot showing RyR, SERCA2a, triadin, and calsequestrin. Increasing amounts (5-20 g) of tissue homogenates from six individual wild-type (WT) and heterozygous (HET) hearts were subjected to SDSpolyacrylamide gel electrophoresis and immunoblotting as described previously. To measure SERCA2a level, 1-, 2-, 4-, and 8-g homogenates were loaded, and the signals were quantitated to confirm the decreased expression of SERCA2 protein in heterozygous hearts. ␣-Actin was used as internal control.
SERCA pumps for Ca 2ϩ (4, 31), was not significantly altered in heterozygous mice compared with wild-type mice. Our results from Ca 2ϩ uptake measurements provided additional supporting evidence that the apparent pump affinity for Ca 2ϩ (K 0.5 ) is unchanged in heterozygous hearts (5). This result may suggest that changes in the SERCA pump level can affect PLB protein level. Although our earlier studies showed that PLB mRNA levels were unchanged in heterozygous hearts (5); here we demonstrate a clear reduction in PLB protein level. Thus, our data indicate that PLB protein level can be regulated posttranscriptionally, possibly by alterations in the rate of protein translation.
In addition to a reduction in total PLB protein level, we found that the ratio of oligomeric PLB (PLB H ) to monomeric PLB (PLB L ) increased ϳ1.4-fold in heterozygous mice. This shift in expression to predominantly more PLB H in heterozygous hearts may be another important compensatory alteration, because PLB L has been shown to be a more effective inhibitor of the SERCA pump than PLB H (39 -41). Additionally, equilibrium between these two states may be regulated by PLB phosphorylation status (39) and SERCA pump level (42). A recent study using fluorescence energy transfer to study reconstitution of PLB with purified Ca 2ϩ -ATPase provided in vitro evidence that an increased SERCA:PLB protein ratio favors the shift from PLB oligomer to PLB monomer, thereby altering the interaction of PLB to SERCA (42). Our data demonstrate for the first time that decreased SERCA pump level can influence the oligomer:monomer ratio, suggesting that an optimal PLB H :SERCA2 interaction is important for the regulation of SERCA pump activity in vivo.
Another important finding of this study is that the basal phosphorylation status of PLB (in vivo) in SERCA2 heterozygous hearts is significantly enhanced at both Ser-16 and Thr-17 despite decreased PLB protein level. It has been well established that phosphorylation of PLB relieves the inhibition and accelerates SERCA pump activity (4,31,32). Phosphorylation of PLB occurs at both Ser-16 and Thr-17 sites by the action of protein kinase A and CaM kinase II, respectively, in response to ␤-adrenergic stimulation (4,43). Therefore, one plausible mechanism for the increased PLB phosphorylation is an increase in basal sympathetic drive in heterozygous mice leading to increased activation of protein kinase A as well as CaM kinase II. Alternatively, SR Ca 2ϩ load might itself directly regulate PLB phosphorylation, as suggested by the recent study by Bhogal and Colyer (44). They have shown that depletion of Ca 2ϩ from cardiac SR stimulates an SR intrinsic protein kinase to phosphorylate PLB at Ser-16, promoting the refilling of SR Ca 2ϩ stores through increased SERCA pump activity.
Taken together, our data suggest that a decrease in PLB protein level, a shift from PLB monomer to oligomer, and an increase in PLB phosphorylation status work concomitantly to enhance SERCA pump activity. As a result, SERCA pump activity is stimulated in heterozygous hearts to partially compensate for the decrease in SERCA pump level and Ca 2ϩ uptake function.
The Up-regulation of Na ϩ -Ca 2ϩ Exchanger May Play a Compensatory Role in SERCA2 Heterozygous Hearts-It is known that the sarcolemmal Na ϩ -Ca 2ϩ exchanger is an important Ca 2ϩ extrusion mechanism that contributes to cardiac relaxation (35). Therefore, reduced levels of SERCA2 may increase the dependence of SERCA2 heterozygous myocytes on NCX activity to maintain a normal rate of Ca 2ϩ removal. Consistent with this hypothesis, we found increases in the Na ϩ -Ca 2ϩ exchanger protein level, as well as increased exchanger current density in heterozygous hearts, suggesting that it plays a compensatory function in Ca 2ϩ handling. Decreased SERCA2 levels have been associated with increased expression of NCX in a number of different studies. For example, during heart development as well as in heart failure, Na ϩ -Ca 2ϩ exchanger expression levels are up-regulated, whereas SERCA pump levels are down-regulated (45)(46)(47). Similar results were observed in a recent study using hypothyroid mice (10). These data, taken together, suggest that the increased Na ϩ -Ca 2ϩ exchanger expression is a compensatory response to the reduction in SERCA pump level.
The Na ϩ -Ca 2ϩ exchanger has been shown to transport Ca 2ϩ in either direction and, hence, promote Ca 2ϩ entry as well as Ca 2ϩ extrusion (6). If Na ϩ -Ca 2ϩ exchanger activity was increased only in the forward mode, which promotes Ca 2ϩ extrusion, it might lead to a further decrease in SR Ca 2ϩ load. Therefore, we hypothesized that the reverse mode of Na ϩ -Ca 2ϩ exchanger is also up-regulated to maintain the balance between Ca 2ϩ extrusion and Ca 2ϩ entry. In this study, we found that both forward and reverse mode Na ϩ -Ca 2ϩ exchanger currents are increased in SERCA2 heterozygous hearts. There is evidence showing that the enhanced Ca 2ϩ entry via the reverse mode of the Na ϩ -Ca 2ϩ exchanger can trigger SR Ca 2ϩ release (48) and may provide inotropic support for the myocytes (49,50). Our data suggest that an increase in Na ϩ -Ca 2ϩ exchanger expression/activity in forward mode and reverse mode may stimulate Ca 2ϩ extrusion as well as Ca 2ϩ entry to maintain intracellular Ca 2ϩ homeostasis.
In conclusion, our studies show that a decrease in SERCA pump level can result in alterations in Ca 2ϩ homeostasis and a decrease in myocyte contractility. Although several compensatory mechanisms are activated in SERCA2 heterozygous hearts to maintain Ca 2ϩ homeostasis, the net effect is still a deficit in cardiac function, demonstrating that SERCA pump level is a critical determinant of cardiac contractility. FIG. 7. Scatchard plot analysis of ryanodine binding in homogenates from wild-type and heterozygous mouse hearts. Specific [ 3 H]ryanodine binding to cardiac homogenates was measured as described under "Experimental Procedures." Data were obtained from six individual hearts from wild-type (q) and SERCA2 heterozygous (E) mice and were fitted with a single class of binding sites.